Meta just launched an AI that can actually run your life online — here's what Muse can do

(Image credit: Future) Meta has launched Muse, a new personal AI agent designed to work across the apps and services you already use. Unlike a chatbot that gives you an answer and waits for another prompt, Muse can open a browser, fill out forms, send emails, book travel, make purchases and continue working on a task after you close the app.You can even give it a bigger goal and let it figure out the steps required to get there, which makes Muse one of the clearest examples of where consumer AI appears to be headed.But there’s an obvious tradeoff: For Muse to become genuinely useful, you have to give it access to a remarkable amount of your digital life.Muse isn't just another chatbotMeta describes Muse as a “personal AI agent,” which is similar to what we've seen from Gemini and ChatGPT recently. Like the AI agents from Meta competitor's Muse is designed to do more.According to Meta, the agent can open websites, fill out forms and negotiate on your behalf. It can book travel, send emails and make purchases, checking back with you when it needs approval for certain actions.It can also keep working after you close the app, which opens the door for much longer-running tasks. This could include tasks such as trying to lower a bill, adapting a training plan when your schedule changes or even selling a car, according to Meta's examples in the company press release.Muse can also remember information you've previously shared and use it later. Meta says it could notice a recipe Reel you saved on Instagram, turn it into a grocery list and remember your friends' dietary restrictions when helping organize a dinner party.Sign up to the Tom's AI Guide weekly newsletter summing up all the biggest AI news you need to know. Plus, analysis from our AI editors and tips on how to use the latest AI tools!In other words, the goal isn't necessarily for you to keep prompting Muse. It's for Muse to learn enough about your life that it can start anticipating what you might need.It can spend money for you One of the biggest differences between an AI assistant and an AI agent is what happens when money gets involved. Muse can make purchases using Stripe's Link payment system. Instead of giving the agent your actual card number, Link generates a one-time-use card for the transaction.Meta says Muse is also the first AI agent covered by Link's purchase protections, which can include things such as no-fee returns and protection for eligible lost or damaged items.Shop Pay support is expected to arrive later, while Meta says integration with 1Password is also coming so Muse can use credentials stored in a user's password manager.That could make tasks such as shopping considerably more convenient. It also raises the stakes considerably if an agent misunderstands what you asked it to do. (Image credit: Shutterstock)Meta clearly knows that giving an autonomous AI access to email, accounts and payments creates an entirely different security problem. Muse therefore runs inside what Meta calls a Secure VM — essentially its own isolated computer in the cloud.There's also another AI agent running alongside it called Sentinel, separated from Muse at the system level and reviews what the agent is trying to do before its actions reach the internet. Meta says Muse also asks for permission before sensitive actions such as sending an email or making a purchase.Users can decide which services Muse connects to and what permissions it receives. Giving Muse access to email, for example, doesn't necessarily mean allowing it to send messages.There's also an audit trail showing what the agent has done and what it plans to do. Meta says Muse doesn't see passwords or payment information directly. Credentials are kept in secure storage and can be used by the agent without revealing the underlying information to it.Later this year, Meta plans to introduce what it calls Muse Confidential VM, which will encrypt the entire virtual machine with a key held by the user. Meta says that would prevent even the company itself from accessing the data and conversations stored inside.Safeguards and trust All of those safeguards matter because Muse potentially has access to some of the most sensitive parts of someone's digital life. And Meta's own testing shows why autonomous agents remain difficult to secure.Reuters reports that Meta originally planned to release Muse earlier this year but delayed the launch while it worked on security.Internal testing reportedly produced mixed results. In one particularly concerning example, an employee said the agent managed to get around safeguards and expose personal iCloud photos after being asked to identify toys visible in pictures from a child's birthday party.Other employees reported reliability problems, including an agent that stopped refreshing a webpage while monitoring hard-to-find tickets and sometimes silently ignored errors.Meta told Reuters that it had done enough additional security work to reach the threshold required to release Muse, although the company acknowledged that mistakes remain possible.Final thoughts We've now entered a post-chatbot era. There are already plenty of AI agents capable of using computers, browsers and other tools. What's potentially significant about Muse is who Meta wants using it.Muse is rolling out in the U.S. on iOS and Android and through the web, and you can also talk to it directly through WhatsApp. Meta says support for its AI glasses is coming soon. There's a free tier, with paid plans available for people who need heavier usage.Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok. More from Tom's GuideAI is turning ordinary photos of kids into sexual abuse imagery — here's how to protect your familyI asked ChatGPT, Gemini, Claude and Perplexity to rank the 10 jobs most likely to be replaced by AI — here's what they all agreed onGemini just got a 'floating bubble' multitasking upgrade that's a game changer on Android phones — here's how to use it Amanda Caswell is the AI Editor at Tom's Guide and one of today’s leading voices in AI and technology. A celebrated contributor to various news outlets, her sharp insights and relatable storytelling have earned her a loyal readership. Amanda’s work has been recognized with prestigious honors, including outstanding contribution to media.Known for her ability to bring clarity to even the most complex topics, Amanda seamlessly blends innovation and creativity, inspiring readers to embrace the power of AI and emerging technologies. As a certified prompt engineer, she continues to push the boundaries of how humans and AI can work together.Beyond her journalism career, Amanda is a long-distance runner and mom of three. She lives in New Jersey.